with intravitreal administration of 9-cis-retinal in rpe65 ... · with intravitreal administration...

ONLINE FIRST

LABORATORY SCIENCES

Improvement of Visual PerformanceWith Intravitreal Administrationof 9-cis-Retinal in Rpe65-Mutant DogsPatricia M. Gearhart, DVM, PhD; Chris Gearhart, PhD;Debra A. Thompson, PhD; Simon M. Petersen-Jones, DVetMed, PhD

Objective: To determine the efficacy of intravitreal ad-ministration of 9-cis-retinal in restoring visual functionin Rpe65-mutant dogs.

Methods: Intravitreal injection of 9-cis-retinal was ad-ministered in 1 eye of 7 Rpe65−/− dogs at a range of ages.Electroretinogram analysis and testing of visual perfor-mance was used to evaluate outcomes after a single injec-tion and in 2 dogs after a second injection in the same eye.

Results: In 5 of 7 injected dogs, 9-cis-retinal injectionresulted in increased rod electroretinogram responses andimproved functional vision. Three injected dogs exhib-ited increased 33-Hz flicker amplitudes characteristic ofcone-mediated responses. Electroretinogram improve-ment was no longer evident by week 10 postinjection in1 dog monitored over time. A second injection of 9-cis-

retinal was performed in the same eye of 2 of the 7 dogsand also resulted in rescue of visual function.

Conclusion: Our findings establish that 9-cis-retinoidtherapy can restore visual function in a canine model ofhuman disease resulting from RPE65 mutations.

Clinical Relevance: These positive proof-of-principleresults provide support for the development of intravit-real devices for sustained delivery of 9-cis-retinal as atherapy for conditions resulting from failure of thevisual cycle.

Arch Ophthalmol. 2010;128(11):1442-1448.Published online September 13, 2010.doi:10.1001/archophthalmol.2010.210

L EBER CONGENITAL AMAURO-sis is a genetically heterog-enous condition that is an im-portant cause of severe visualdisability in children and

young adults.1 Mutations in RPE65 are re-sponsible for approximately 6% of autoso-mal recessive Leber congenital amaurosis,often referred to as Leber congenital amau-rosis type 2, with patients progressing to le-gal blindness in early adulthood.2

RPE65 is an essential component of thevisual cycle where it acts as a retinoid isom-erase to convert esters of vitamin A storedin the retinal pigment epithelium to 11-cis-retinol, which is then oxidized to 11-cis-retinal and transported to the retinawhere it combines with rhodopsin andcone opsins. Disruption of the Rpe65 genein knockout mice results in a lack of 11-cis-retinal and electroretinogram (ERG) re-sponses, downregulation of rhodopsin pro-tein, and accumulation of retinyl esters inthe retinal pigment epithelium.3

Groundbreaking research demonstrat-ing the efficacy of gene therapy for inher-ited retinal degeneration centered aroundthe Briard breed of dog that carries a natu-rally occurring 4–base pair deletion inRpe65.4 Gene therapy was effective in res-cuing the disease phenotype.5-8 These stud-ies formed a critical foundation for work

that has now progressed to phase 1/2 clini-cal trials of gene therapy in human pa-tients with Leber congenital amaurosis,with the first reported outcomes show-ing promising results.9-11

As an alternative approach, studies withRpe65 knockout mice have shown that ret-inoid supplementation strategies involv-ing oral delivery of 9-cis-retinal, and intra-peritoneal injection of 11-cis-retinal, areeffective in reconstituting active visual pig-ments and visual responses.12,13 Further-more, a combination of retinoid and genetherapy was shown to be complementaryand resulted in improved rescue com-pared with each treatment alone in mice

Video available online atwww.archophthalmol.com

For editorial commentsee page 1483

Author Affiliations: Departmentof Small Animal ClinicalSciences, College of VeterinaryMedicine, Michigan StateUniversity, East Lansing(Drs P. M. Gearhart, C. Gearhart,and Petersen-Jones), andDepartment of Ophthalmologyand Visual Sciences, Departmentof Biological Chemistry,University of Michigan MedicalSchool, Ann Arbor(Dr Thompson).

(REPRINTED) ARCH OPHTHALMOL / VOL 128 (NO. 11), NOV 2010 WWW.ARCHOPHTHALMOL.COM1442

©2010 American Medical Association. All rights reserved.

Downloaded From: on 07/12/2018

with a failure of the visual cycle due to a lack of lecithinretinol acyltransferase.14 Oral retinoid therapy has en-tered phase 1b clinical trials, with early reports suggest-ing evidence of efficacy in patients with lecithin retinol ac-yltransferasemutations.15 Long-termadministrationof9-cis-retinyl acetate in aging mice improved ERG responses anddark adaptation, leading to the suggestion that such thera-pies may protect against age-related retinal dysfunction.16

Despite its successes, gene therapy is unlikely to be thefirst-line treatment for all genetic defects affecting visualcycle function, especially those associated with disease thatis less severe and of later onset (eg, mutations in RDH5 andRLBP117,18). In addition, adjunct therapies that augmentother treatment paradigms are likely to be needed in spe-cific circumstances, eg, in cases involving gene inactiva-tion and where repeated injections of viral vectors mightnot be advisable. Because of these issues and other consid-erations relative to the development of gene therapy ap-proaches,19 the purpose of our proof-of-principle study wasto determine if retinoid therapy could rescue the Rpe65-mutant dog phenotype. We now show that intravitreal in-jection of 9-cis-retinal can result in markedly improved ERGresponses and visual performance in the Rpe65-mutant dogthat can be reinitiated by repeated injection.

METHODS

ANIMALS

Seven Rpe65−/− dogs were used in this study (Table 1). Allprocedures were approved by the Institutional Animal Care andUse Committee. Prior to 9-cis-retinal injections, baseline vi-

sion testing of each eye and bilateral ERG recordings were made.Following 9-cis-retinal injections, dogs were kept in the darkwith dim red light used during cleaning, care, and mainte-nance. In 2 dogs (dogs 5 and 6), a second injection of 9-cis-retinal was made into the same eye. Ophthalmoscopically nor-mal Rpe65�/� crossbred dogs of similar ages were used ascontrols for normal ERG responses.

ANESTHESIA

Dogs were anesthetized by premedication with acepromazine ma-leate (0.1-0.3 mg/kg) intramuscularly, induction with thiopen-tal sodium (6-12 mg/kg intravenously), and maintenance withisoflurane (1% to 2% delivered in oxygen via an endotracheal tube).

9-cis-RETINAL PREPARATION AND INJECTION

9-cis-Retinal (R5754, Lot 026K1125; Sigma Chemical, St Louis,Missouri) was resuspended in ethanol at 344mM and puritywas evaluated by normal-phase high-performance liquid chro-matography. The concentration was calculated from the 373-nmabsorbance maximum and the extinction coefficient 36 068M−1

cm−1 (in ethanol).20 For injections, 4 µL of 9-cis-retinal stocksolution (391 µg) was mixed with lactated Ringer solution tomake 100 µL. Four-microliter ethanol in 100-µL lactated Ringersolution was used as a control vehicle injection. All retinoidmanipulation was performed under dim red light.

9-cis-RETINAL INJECTION

Standard Intravitreal Injection

In the early part of the study (dogs 1-4), 9-cis-retinal was de-livered by a conventional intravitreal injection. With the dogunder general anesthesia and in lateral recumbency, the ocu-

Table 1. Details of the Dogs Used in the Study and ERG Outcomes Using a Measure of Rod and Cone Functiona

Dog No./Sex/Age at

Injection, wk Eye TreatmentVision Testing

Technique

PretreatmentDark-Adapted b-Wave

Amplitude, µV,at 0 log cdS/m2

1 wk PostinjectionDark-Adapted b-Wave

Amplitude, µVat 0 log cdS/m2

Pretreatment33-Hz FlickerAmplitude, µV

1 wk Posttreatment33-Hz FlickerAmplitude, µV

1/M/28 OD Intravitreal 9-cis-retinal Obstacle course 1.5 36 4 8.5OS Intravitreal vehicle Obstacle course 3 2 4 3.7

2/M/31 OD Intravitreal vehicle Obstacle course 0 0 2.5 2.6OS Intravitreal 9-cis-retinal Obstacle course 0 20 1.9 6.1

3/F/97 OD Intravitreal vehicle Obstacle course 0 0 2 2OS Intravitreal 9-cis-retinal Obstacle course 0 0 2 2

4/F/12 OD Intravitreal vehicle Obstacle course 0 0 0 0OS Intravitreal 9-cis-retinal Obstacle course 0 4.6 0 0

5/F/42 OD Preretinal 9-cis-retinal Vision testing box 1 21 0 7.6OS Preretinal vehicle Vision testing box 2 0 0 0

6/F/62 OD Preretinal 9-cis-retinal Vision testing box 0 28 2 2OS Preretinal vehicle Vision testing box 0 0 2 2

7/M/90 OD Preretinal vehicle NT 1 1 2 2OS Preretinal 9-cis-retinal NT 1 12.7 2 4.2

Second Injection5/F/71 OD Preretinal 9-cis-retinal Vision testing box 0 3.9 0 4

OS Preretinal vehicle Vision testing box 0 0 0 06/F/82 OD Preretinal 9-cis-retinal Vision testing box 2 17.6 1 12

OS Preretinal vehicle Vision testing box 0 0 0 0Normal control

dogs (n=9),mean (SD)b

46 (18) 21 (9)

Abbreviations: ERG, electroretinogram; NT, not tested.aSee text for details.bShows the ERG results from 18 Rpe65�/� eyes of dogs of similar breeding.




lar surface was prepared for injection using dilute (1 in 50) po-vidone-iodine solution. The dorsolateral bulbar conjunctiva wasgrasped with forceps and the globe was rotated ventromedi-ally. A 27-gauge needle on the Hamilton syringe containing thematerial for injection was inserted approximately 7 mm pos-terior to the sclera, angled posteriorly so as to miss the lens,and the material was injected.

Preretinal 9-cis-Retinal Injection

For the latter part of the study, we developed and used an in-jection technique to deposit the 9-cis-retinal on the surface ofthe retina. Under general anesthesia, the dog was positioned indorsal recumbency with the neck flexed to achieve a horizontalcornea when the eye was in a primary gaze position. The eye wasprepared with dilute (1 in 50) povidone-iodine solution, an eye-lid speculum was fitted, and the globe was positioned using con-junctival stay sutures of 4-0 silk. A subretinal injector (Reti-naJect; SurModics Inc, Irvine, California) was used for thepreretinal injections. The injector was inserted through the parsplana region and advanced toward the retina under direct visu-alization through an operating microscope (Zeiss OpM1 Oper-ating Microscope; Carl Zeiss, Inc, Thornwood, New York) usinga Machemer Magnifying Vitrectomy contact lens (Ocular In-struments Inc, Bellevue, Washington). The 39-gauge extend-able cannula of the RetinaJect was advanced until it was at theretinal surface. At that point, the room lights and the micro-scope light were turned off and the injection was made.

ERG RECORDINGS

Electroretinograms were performed in dog 1 at 1 hour after in-jection and then every week for 4 weeks. In dog 2, ERGs wereperformed at 1 hour, then every week for 8 weeks, and thenevery 2 weeks until 18 weeks. In all other dogs, ERGs were per-formed 1 week after injection. Dark-adapted intensity series;5-Hz rod flicker, light-adapted intensity series; and 33-Hz coneflicker ERGs were recorded as previously described21 except thatERG-Jet lens electrodes (The Electrode Store, Enumclaw, Wash-ington) were used.

ERG DATA ANALYSIS

The a- and b-wave amplitudes were measured as previously de-scribed.21 Electroretinogram amplitudes were plotted as a func-tion of light stimulus. For the flicker responses, amplitude(trough to peak) and implicit times (flash onset to peak am-plitude) were measured.

VISION TESTING

This was performed 1 week after treatment. For dogs 1 through4, a subjective assessment of vision was performed by observingtreated animals moving through an obstacle course of traffic coneswith either the treated or untreated eye patched. The dog wasfilmed using a Sony TRV85 video camera with Nightshot (Sony,San Diego, California). Rpe65−/− dogs maintain some vision inbright lighting conditions for the first few years of life. We there-fore assessed vision at a lighting level at which all untreatedRpe65−/− dogs were unable to successfully negotiate the obstaclecourse. The video recording of the vision testing sessions was as-sessed by 2 independent observers unaware of the treatment sta-tus of the dog.

For dogs 5 and 6, objective vision testing was performedaccording to a method that we have previously described.22

Briefly, this uses a testing device consisting of a chamber with4 exit tunnels. One random tunnel was open for each run of

the test. The first choice of exit tunnel and the time taken toexit were recorded. Performance was analyzed under 3 light-ing intensities (35-45 [room light], 1-1.5, and 0.02-0.04 cd/m2).Each eye was tested separately by masking each eye in turn.

STATISTICAL ANALYSIS

Paired-sample t tests were used to test for differences in ERG am-plitudes between 9-cis-retinal–treated eyes and vehicle-treated eyesand between preinjection and postinjection eyes, as well to as-sess for any difference between first and second injection for dogs5 and 6. For vision testing outcomes, mean time to exit, and meannumber of correct exits, paired-sample t tests were again used totest for differences between 9-cis-retinal–treated eyes and vehicle-treated eyes and between preinjection and postinjection eye out-comes. Paired-sample t tests were chosen instead of more com-plex tests because of the relatively low sample size of this study.Data were considered significant at P� .05.

RESULTS

ERG RESPONSES OF TREATEDRpe65-MUTANT DOGS

In 5 of the 7 Rpe65−/− dogs, injection of 9-cis-retinal re-sulted in restoration of dark-adapted ERG responses ofnormal shape and with a reduced (improved) responsethreshold (Figure 1). In the 2 dogs that had an ERGrecorded at 1 hour after 9-cis-retinal injection, only 1showed an ERG improvement (data not shown) at thatpoint. However, both dogs showed an ERG improve-ment at 1 week after treatment (this was the next pointat which an ERG was recorded). The mean dark-adapted intensity response curve of the 5 dogs with im-proved ERG tracings (Figure 2) shows a lowering ofthreshold for a-wave of about 1.5 log units and for b-wave of about 2.5 log units and a significant increase inamplitude compared with the vehicle-injected eye. Al-though significantly improved, ERG amplitudes re-mained significantly decreased compared with the nor-mal control dogs (Figure 2).

Of the 4 dogs injected by a standard intravitreal in-jection (dogs 1-4), 2 (dogs 1 and 2) showed an ERG res-cue and 2 (dogs 3 and 4) showed no rescue. We consid-ered variability of positioning of the 9-cis-retinal injectionwithin the vitreous as a possible cause for these differ-ing results. To eliminate this possibility, we developed atechnique to place the injected material at the retinal sur-face. Dogs 5, 6, and 7, and the second injections in dogs5 and 6, were performed using this technique and allshowed ERG rescue at 1 week after injection.

Dogs 1 and 2 had weekly ERGs performed for the first4 weeks. There was a progressive decline in ERG rescue(Figure 1 shows the comparison between rescue at 1 and4 weeks posttreatment in dog 1). Dog 2 was followed upfor a longer period by ERG, and the 9-cis-retinal– andvehicle-injected eyes had comparable ERG responses by10 weeks following treatment (data not shown).

Rod responses were evaluated using the dark-adapted b-wave amplitude at 0 log cdS/m2, a flash inten-sity at which untreated Rpe65−/− canine eyes had no re-sponse, or at most a very small response (�5 µV). The




amplitudes at this flash intensity pretreatment and post-treatment are shown in Table 1.

To evaluate cone responses, the amplitudes of 33-Hzflicker responses were analyzed. By this measure, coneERG rescue was not as dramatic as that of rod-mediatedresponses (Table 1). Three of the 7 dogs had an improve-ment in 33-Hz flicker amplitudes above the levels seenin all Rpe65−/− dogs preinjection and in all eyes follow-ing vehicle injection.

SECOND ADMINISTRATION OF 9-cis-RETINAL

Dogs 5 and 6 received a second administration of 9-cis-retinal in the same eye. There was an improvement inERG responses at 1 week following treatment (Figure 3and Table 1). The dark-adapted ERG after the secondtreatment had a slightly higher response threshold com-pared with that after the first injection. However, therewas no significant difference in the a-wave amplitudesbetween the 2 treatments. The b-wave amplitudes fol-lowing the second treatment were significantly lower at2 of the dimmer flash intensities (−2.0 and −0.8 log cdS/m2; P=.04 and .04, respectively) compared with the first.There was no significant difference between treatmentsin b-wave amplitudes at other flash intensities (P valuesranged from .05 to .47). The treated eyes had signifi-cantly larger a- and b-wave amplitudes at all flash inten-sities compared with the vehicle-injected eyes (P valuesranged from .049 to .001).

VISION ASSESSMENTIN TREATED Rpe65−/− DOGS

Independent observers blinded to the treatment status ofthe dogs found that both dogs 1 and 2 negotiated the ob-stacle course in dim lighting conditions with fewer colli-sions with obstacles and more rapidly when the 9-cis-retinal–injected eye was uncovered compared with when

the vehicle-injected eye was uncovered (video, http://www.archophthalmol.com). There was no difference in visualperformance detected between 9-cis-retinal–treated and ve-hicle-treated eyes of dogs 3 and 4. These results correlatedwith the ERG findings. To provide data for statistical analy-sis of vision testing, we developed an objective vision testas described earlier.22 For dogs tested with this technique(Table2), at the 2 lower–light intensity ranges (1-1.5 cd/m2

and 0.02-0.04 cd/m2), there was a statistically significantimprovement in both mean selection of the correct exit tun-nel (P=.003 and P=.03, respectively) and the mean timeto exit the device (P=.008 and P=.02, respectively) withthe treated eye uncovered compared with the vehicle-treated eye uncovered. In normal room lighting (35-45 cd/m2), there was no significant difference in mean correctchoice and mean time to exit (P=.19 and P=.15).

COMMENT

Our findings establish that intravitreal administration of9-cis-retinal is effective in restoring useful visual functionin Rpe65-deficient dogs. Using optimized injection proto-cols, we consistently obtained improved responses that wereattributable to rod photoreceptor function. These resultsprovide strong support for the potential of retinoid deliv-ery as a possible treatment for RPE65 loss of function.

Previous studies of Rpe65−/− mice established the ef-fectiveness of systemic administration (both oral and in-travenous) of 9-cis-retinal for restoring ERG activity13,23

and repeated oral administration was also effective.24 In-traperitoneal injection of 11-cis-retinal was also shownto improve the ERG function of Rpe65−/− mice12 and whenadministered to young animals, corrected the subcellu-lar mislocalization of cone opsin resulting from chro-mophore loss and associated with cone cell death.25 How-ever, the doses of retinoid required for rescue wereextremely high, leading to concerns over potential toxic

Pretreatment

Right LeftDark adapted

RightRod flicker

LeftCone flickerLight adapted

1 wk

Pretreatment

1 wk

4 wk4 wk

Pretreatment

1 wk

4 wk

A B

Figure 1. Comparison of electroretinogram responses from pretreatment and then 9-cis-retinal–injected and vehicle-injected eyes of an Rpe65−/− dog (dog 1).A, Dark-adapted and light-adapted electroretinogram intensity response series from pretreatment, 1 week after treatment, and 4 weeks after treatment. The righteye was injected with 9-cis-retinal and the left eye received vehicle. The vertical line across the tracings indicates the timing of the flash. Vertical size bar fordark-adapted series=50 µV; for light-adapted series=20 µV. Horizontal bars=50 milliseconds. Dark-adapted flash intensities range from −2.79 to 2.4 log cdS/m2.Light-adapted flash intensities range from −0.2 to 2.8 log cdS/m2. B, Rod (dark-adapted 5-Hz series, −1.6 log cdS/m2) and cone (light-adapted to 30 cd/m2 33-Hzseries, 0.39 log cdS/m2) flicker preinjection and 1 and 4 weeks after injection. The upper tracings are from the right eye (9-cis-retinal treated) and the lowertracings, from the left eye (vehicle injected). Vertical bars are 5 µV. Horizontal bars are 500 milliseconds for 5-Hz flicker and 50 milliseconds for 33-Hz flicker.Dashed vertical lines indicate timing of flashes.




effects26-28 and motivating us to develop an effective ap-proach for limited and local delivery. The dose used inthe current study (about 400 µg of 9-cis-retinal per eye)translates to about 20 µg/kg of body weight, which is con-siderably lower than the dose used for systemic admin-istration in mice, which was between 250 and 2500 µgper mouse, which is in the region of 1.6 to 16�104 µg/kgof body weight.13,23

Electroretinogram analysis of Rpe65-mutant dogstreated with 9-cis-retinal showed increased dark-adapted ERG amplitudes in 5 of 7 treated dogs and coneflicker responses in 3 of 7 treated dogs. Rod responseswere also compared at a flash intensity (0 log cdS/m2)that is less than the response threshold for most un-treated Rpe65-mutant dogs (although in a few dogs a b-wave of �5 µV could be recorded). The lack of ERG at

this intensity is similar to that reported by others.5 Vi-sion testing correlated with the recovery of dark-adapted ERG responses. Cone function was assessed byanalyzing 33-Hz flicker amplitudes rather than the single-flash light-adapted responses. The flicker responses wereselected because the dramatically reduced rod sensitiv-ity in RPE65-deficient animals29 could mean that nor-mal rod-suppressing background illumination might notcompletely eliminate responses from untreated rods. Re-sponses from these rods might contaminate the light-adapted single-flash responses. However, rods (treatedor untreated) are not able to recover rapidly enough tocontaminate 33-Hz flicker responses. Convincing im-provements in 33-Hz flicker responses in 3 of the 7 treateddogs provide strong evidence of cone rescue. Our re-sults suggest that a fraction of the cone cell population

1000

10

100

1

0.1– 3 – 2 – 1 10 2 3

Intensity, log cdS/m2

a-W

ave

Ampl

itude

, µV

A

1000

100

10

1– 4 – 3 – 2 – 1 10 2 3


b-W

ave

Ampl

itude

, µV

BRpe65 +/+ eyes

Vehicle-injected eyes9-cis -Retinal–injected eyes

Figure 2. Dark-adapted mean a- and b-wave electroretinogram intensity response plots (log:log scale) for normal dogs of similar breeding (n=9), the treated dogswith rescue (n=5), and the vehicle-injected eyes of the same dogs (n=5). Error bars are �1 SD. The mean a-wave (A) and mean b-wave (B) amplitudes for thetreated eyes are significantly greater than those of the vehicle-injected eyes at each flash intensity (P� .05).

100

1

10

0.1– 2 – 1 10 2 3


a-W

ave

Ampl

itude

, µV

A

100

10

1

– 4 – 3 – 2 – 1 10 2 3Intensity, log cdS/m2

b-W

ave

Ampl

itude

, µV

BFirst injection

VehicleSecond injection

Figure 3. Result of the second 9-cis-retinal injection in the same eye for dogs 5 and 6. The mean (SD) a-wave (A) and b-wave (B) dark-adapted intensity response ampli-tudes are shown. In each graph, the solid line is the mean result from the first 9-cis-retinal injection, the long dashed line is the mean result from the second 9-cis-retinalinjection, and the short dashed lines are the mean results from the vehicle-injected eyes. The second injection was performed 29 weeks after the first for dog 5 and 20weeks later for dog 6. Electroretinograms before the second injection showed that response had returned to baseline following initial rescue by the first injection.




remains responsive to intravitreal retinoid in Rpe65−/−dogs up to at least 82 weeks of age (dog 6, second injec-tion). This observation is noteworthy in the context ofprevious studies showing that cone cell death begins inRpe65−/− mice at a few weeks of age.30,31 Early cone celldeath is also evident by in vivo imaging of patients withRPE65-null mutations, including young children32-34; how-ever, a subpopulation of cone cells appears to retain nor-mal structure and function for longer.

The visual testing results correlated closely with theERG findings. We have previously reported that in brightroom-lighting conditions young Rpe65-deficient dogs havereasonable vision, allowing them to exit from our visiontesting device22 and to negotiate an obstacle course (datanot shown). In fact, the condition in the Swedish Briardwas first described as a congenital stationary night blind-ness.35 We have shown that as the room lighting is low-ered from normal levels they start to have problems see-ing and at lower lighting levels are unable to see the exitfrom the vision testing device.22 Following successful treat-ment (as assessed by ERG outcome), the visual perfor-mance was significantly improved at lower lighting con-ditions (Table 2 and video).

The failure of intravitreal injection to provide evidenceof rescue (by ERG or vision testing) in 2 dogs may reflectthe initial injection technique used, in which there waspotential for considerable variation of site of 9-cis-retinaldeposition. This could vary from close to the retinal sur-face to midvitreous. 9-cis-Retinal is very hydrophobic, andtherefore, there is probably limited diffusion from the in-jectionsite.Wetheorized thatvariability in thesiteofdepo-sition of 9-cis-retinal could have accounted for the vari-ability in whether rescue was achieved. Development ofa technique to accurately place the injected material at theretinal surface resulted inERG-detectable rescue inall eyesinjected in this fashion (1 eye injected once and 2 eyes in-jected twice in dogs 5-7). Additional experiments to lookat diffusion of the retinoid within the vitreous body wouldbe required to test our theory for the reason for failure oftreatment in dogs 3 and 4. Clearly, the method of injec-tion, injection site, dose given, and diffusion character-isticsoftheretinoidhavethepotential tosignificantlyimpactoutcomes and will form the basis for future refinementsof the approach.

The improvements in ERG function obtained follow-ing a single injection of 9-cis-retinal were lost by 10 weeksfollowing injection indicating that repeated administra-

tion or development of a sustained delivery device will berequired. Importantly, we found that repeated 9-cis-retinal injection into a previously treated eye resulted inrescue as assessed by ERG and vision testing. The dark-adapted ERG a- and b-wave intensity response curves weresimilar between the first and second treatments, but theredid appear to be slight differences at lower flash intensi-ties with an increase in response threshold and a signifi-cant reduction in b-wave amplitude at 2 of the low-intensity flashes. The cone ERG results were differentbetween the injections, with dog 5 having a lower coneERG recordable at the second injection compared with thefirst, whereas dog 6 had no detectable cone ERG improve-ment at the first injection but a good improvement afterthe second injection. It is possible that the proximity ofthe injection to the regions of highest cone density (areacentralis and visual streak36) coupled with limited reti-noid diffusion may have accounted for some of this ap-parent disparity in cone rescue. This warrants further study.The RPE65-deficient phenotype is characterized by a pro-gressive loss of cone photoreceptors, which could con-tribute to the difference in cone responses between the 2points. There appear to be species differences in the rateof cone degeneration, with mice losing cones rapidly inearly life30 whereas humans have an early cone loss buthave a population of cones that are maintained for de-cades.32,33 Cone survival in Rpe65-deficient dogs has notbeen studied in detail, but ultrastructural studies in youngerdogs showed that cone outer segments were better pre-served than rod outer segments,37 and in a single older (7-year-old) dog that was examined, cones were preservedin the central retina and had intact outer segments.38 Inview of the apparently slow cone degeneration in the dogmodel, it seems unlikely that cone degeneration made alarge contribution to the reduced cone ERG rescue seenafter the second injection in dog 6.

The approach reported herein is the first, to our knowl-edge, to show the efficacy of intravitreal administrationof 9-cis-retinal in an RPE65-deficient animal model anddescribes a technique for accurate injection of the reti-noid at the retinal surface. Our long-term goal is the de-velopment of sustained-release delivery of retinoids asan alternative treatment for patients with failure of thevisual cycle.

Submitted for Publication: January 18, 2010; final re-vision received June 3, 2010; accepted June 16, 2010.

Table 2. Vision Testing Results

Light Intensity,cd/m2

Mean

Frequency of Correct Tunnel Choice (7 runs) Time to Exit Device, s

VehicleInjected

9-cis-RetinalTreated

P Value:9-cis-Retinal Treated

vs Vehicle TreatedVehicleInjected

9-cis-RetinalTreated

P Value:9-cis-Retinal Treated

vs Vehicle Treated

35-45 5.8 7 .19 5.7 4.3 .151-1.5 3.8 5.6 .003a 18.9 10.2 .008a

0.02-0.04 2.2 5.2 .03a 20.8 13.0 .02a

aSignificant at P� .05.




Published Online: September 13, 2010. doi:10.1001/archophthalmol.2010.210Correspondence: Simon M. Petersen-Jones, DVetMed,PhD, Department of Small Animal Clinical Sciences,Michigan State University, D-208 Veterinary Medical Cen-ter, East Lansing, MI 48824 ([email protected]).Financial Disclosure: None reported.Funding/Support: Funding was received from GlassenFoundation (Drs Petersen-Jones and P. M. Gearhart), CVMPurebred Dog Endowment Fund (Drs Petersen-Jones andP. M. Gearhart), Michigan Eye Bank (Drs Petersen-Jonesand P. M. Gearhart), Foundation Fighting Blindness (DrThompson), and a Senior Scientific Investigator Award fromResearch to Prevent Blindness (Dr Thompson).Online-Only Material: The video is available at http://www.archophthalmol.com.Additional Contributions: Janice Forcier, BS, NalineeTuntivanich, DVM, PhD, Cheri Johnson, DVM, Lisa Allen,BS, and Christina L. McHenry, PhD, provided expert tech-nical assistance.

REFERENCES

1. Daiger SP, Bowne SJ, Sullivan LS. Perspective on genes and mutations causingretinitis pigmentosa. Arch Ophthalmol. 2007;125(2):151-158.

2. Hanein S, Perrault I, Gerber S, et al. Leber congenital amaurosis: comprehen-sive survey of the genetic heterogeneity, refinement of the clinical definition, andgenotype-phenotype correlations as a strategy for molecular diagnosis. Hum Mutat.2004;23(4):306-317.

3. Redmond TM, Yu S, Lee E, et al. Rpe65 is necessary for production of 11-cis-vitamin A in the retinal visual cycle. Nat Genet. 1998;20(4):344-351.

4. Veske A, Nilsson SE, Narfström K, Gal A. Retinal dystrophy of Swedish briard/briard-beagle dogs is due to a 4-bp deletion in RPE65. Genomics. 1999;57(1):57-61.

5. Acland GM, Aguirre GD, Ray J, et al. Gene therapy restores vision in a caninemodel of childhood blindness. Nat Genet. 2001;28(1):92-95.

6. Jacobs JB, Dell’Osso LF, Hertle RW, Acland GM, Bennett J. Eye movement re-cordings as an effectiveness indicator of gene therapy in RPE65-deficient ca-nines: implications for the ocular motor system. Invest Ophthalmol Vis Sci. 2006;47(7):2865-2875.

7. Narfström K, Katz ML, Ford M, Redmond TM, Rakoczy E, Bragadottir R. In vivogene therapy in young and adult RPE65-/- dogs produces long-term visualimprovement. J Hered. 2003;94(1):31-37.

8. Narfström K, Seeliger M, Lai CM, et al. Morphological aspects related to long-term functional improvement of the retina in the 4 years following rAAV-mediatedgene transfer in the RPE65 null mutation dog. Adv Exp Med Biol. 2008;613:139-146.

9. Bainbridge JW, Smith AJ, Barker SS, et al. Effect of gene therapy on visual func-tion in Leber’s congenital amaurosis. N Engl J Med. 2008;358(21):2231-2239.

10. Maguire AM, Simonelli F, Pierce EA, et al. Safety and efficacy of gene transferfor Leber’s congenital amaurosis. N Engl J Med. 2008;358(21):2240-2248.

11. Hauswirth WW, Aleman TS, Kaushal S, et al. Treatment of Leber congenital am-aurosis due to RPE65 mutations by ocular subretinal injection of adeno-associated virus gene vector: short-term results of a phase I trial. Hum Gene Ther.2008;19(10):979-990.

12. Ablonczy Z, Crouch RK, Goletz PW, et al. 11-cis-Retinal reduces constitutive op-sin phosphorylation and improves quantum catch in retinoid-deficient mouse rodphotoreceptors. J Biol Chem. 2002;277(43):40491-40498.

13. Van Hooser JP, Aleman TS, He YG, et al. Rapid restoration of visual pigmentand function with oral retinoid in a mouse model of childhood blindness. ProcNatl Acad Sci U S A. 2000;97(15):8623-8628.

14. Batten ML, Imanishi Y, Tu DC, et al. Pharmacological and rAAV gene therapyrescue of visual functions in a blind mouse model of Leber congenital amaurosis.PLoS Med. 2005;2(11):e333.

15. Koenekoop RL. An overview of retinal dystrophies: from gene discoveries to newtherapies. Paper presented at: Therapy for Sight-Minisymposium; Associationfor Research in Vision and Ophthalmology Annual Meeting; Fort Lauderdale, FL;May 3, 2010. Program 2028.

16. Maeda T, Maeda A, Leahy P, Saperstein DA, Palczewski K. Effects of long-termadministration of 9-cis-retinyl acetate on visual function in mice. Invest Oph-thalmol Vis Sci. 2009;50(1):322-333.

17. Yamamoto H, Simon A, Eriksson U, Harris E, Berson EL, Dryja TP. Mutations inthe gene encoding 11-cis retinol dehydrogenase cause delayed dark adaptationand fundus albipunctatus. Nat Genet. 1999;22(2):188-191.

18. Morimura H, Berson EL, Dryja TP. Recessive mutations in the RLBP1 gene en-coding cellular retinaldehyde-binding protein in a form of retinitis punctataalbescens. Invest Ophthalmol Vis Sci. 1999;40(5):1000-1004.

19. Kumar-Singh R. Barriers for retinal gene therapy: separating fact from fiction.Vision Res. 2008;48(16):1671-1680.

20. Garwin GG, Saari JC. High-performance liquid chromatography analysis of vi-sual cycle retinoids. Methods Enzymol. 2000;316:313-324.

21. Tuntivanich N, Pittler SJ, Fischer AJ, et al. Characterization of a canine model ofautosomal recessive retinitis pigmentosa due to a PDE6A mutation. Invest Oph-thalmol Vis Sci. 2009;50(2):801-813.

22. Gearhart PM, Gearhart CC, Petersen-Jones SM. A novel method for objective vi-sion testing in canine models of inherited retinal disease. Invest Ophthalmol VisSci. 2008;49(8):3568-3576.

23. Van Hooser JP, Liang Y, Maeda T, et al. Recovery of visual functions in a mousemodel of Leber congenital amaurosis. J Biol Chem. 2002;277(21):19173-19182.

24. Maeda T, Maeda A, Casadesus G, Palczewski K, Margaron P. Evaluation of 9-cis-retinyl acetate therapy in Rpe65-/- mice. Invest Ophthalmol Vis Sci. 2009;50(9):4368-4378.

25. Rohrer B, Lohr HR, Humphries P, Redmond TM, Seeliger MW, Crouch RK.Cone opsin mislocalization in Rpe65-/- mice: a defect that can be corrected by11-cis retinal. Invest Ophthalmol Vis Sci. 2005;46(10):3876-3882.

26. Jacques EA, Buschmann RJ, Layden TJ. The histopathologic progression of vi-tamin A-induced hepatic injury. Gastroenterology. 1979;76(3):599-602.

27. Geubel AP, De Galocsy C, Alves N, Rahier J, Dive C. Liver damage caused bytherapeutic vitamin A administration: estimate of dose-related toxicity in 41 cases.Gastroenterology. 1991;100(6):1701-1709.

28. Sibulesky L, Hayes KC, Pronczuk A, Weigel-DiFranco C, Rosner B, Berson EL.Safety of �7500 RE (�25000 IU) vitamin A daily in adults with retinitis pigmentosa.Am J Clin Nutr. 1999;69(4):656-663.

29. Seeliger MW, Biesalski HK, Wissinger B, et al. Phenotype in retinol deficiencydue to a hereditary defect in retinol binding protein synthesis. Invest Ophthal-mol Vis Sci. 1999;40(1):3-11.

30. Znoiko SL, Rohrer B, Lu K, Lohr HR, Crouch RK, Ma JX. Downregulation of cone-specific gene expression and degeneration of cone photoreceptors in the Rpe65-/-mouse at early ages. Invest Ophthalmol Vis Sci. 2005;46(4):1473-1479.

31. Rohrer B, Goletz P, Znoiko S, et al. Correlation of regenerable opsin with rodERG signal in Rpe65-/- mice during development and aging. Invest OphthalmolVis Sci. 2003;44(1):310-315.

32. Jacobson SG, Aleman TS, Cideciyan AV, et al. Identifying photoreceptors in blindeyes caused by RPE65 mutations: prerequisite for human gene therapy success.Proc Natl Acad Sci U S A. 2005;102(17):6177-6182.

33. Jacobson SG, Aleman TS, Cideciyan AV, et al. Human cone photoreceptor de-pendence on RPE65 isomerase. Proc Natl Acad Sci U S A. 2007;104(38):15123-15128.

34. Maeda T, Cideciyan AV, Maeda A, et al. Loss of cone photoreceptors caused bychromophore depletion is partially prevented by the artificial chromophore pro-drug, 9-cis-retinyl acetate. Hum Mol Genet. 2009;18(12):2277-2287.

35. Narfström K, Wrigstad A, Nilsson SE. The Briard dog: a new animal model ofcongenital stationary night blindness. Br J Ophthalmol. 1989;73(9):750-756.

36. Mowat FM, Petersen-Jones SM, Williamson H, et al. Topographical character-ization of cone photoreceptors and the area centralis of the canine retina. MolVis. 2008;14:2518-2527.

37. Wrigstad A, Nilsson SE, Narfström K. Ultrastructural changes of the retina andthe retinal pigment epithelium in Briard dogs with hereditary congenital nightblindness and partial day blindness. Exp Eye Res. 1992;55(6):805-818.

38. Wrigstad A, Narfström K, Nilsson SE. Slowly progressive changes of the retinaand retinal pigment epithelium in Briard dogs with hereditary retinal dystrophy:a morphological study. Doc Ophthalmol. 1994;87(4):337-354.




with intravitreal administration of 9-cis-retinal in rpe65 ... · with intravitreal administration...

Documents